Perplexity
Perplexity is selling access to other people's models, and now repricing them weekly.
A side-by-side editorial comparison of Dosu and Semantic Kernel — release velocity, themes, recent moves, and the top alternatives to consider.
Dosu is folding agent session logs into the knowledge base it already maintains.
The August Drop turns old agent logs into Dosu knowledge, adds configuration from chat, and surfaces what the agents actually read. It follows Decant by a week — the local tool that parses Claude Code and Codex session logs into per-session cost and activity numbers — so the two releases sit on the same axis from opposite ends: Decant reads the logs on the developer's machine, Dosu now ingests them as a knowledge source. The monthly Drop format continues, with July's removing the waitlist and simplifying Knowledge Cache tools.
Semantic Kernel's releases are now dependency bumps and redirect READMEs pointing users elsewhere.
The .NET and Python packages ship on a steady cadence, but the contents are servicing: SDK and package version bumps, CVE-driven dependency updates, CodeQL suppressions, and HTTP hardening such as disabling automatic redirects in the web plugins. The genuinely functional changes are narrow — a Gemini connector now honoring the configured function choice behavior, an MCP tool approval callback for Azure AI agents shipped as a breaking change, and MCP tools with colliding normalized names being skipped. The latest .NET release removes migrated vector-store providers outright, leaving redirect READMEs behind.
The August Drop turns old agent logs into Dosu knowledge, adds configuration from chat, and surfaces what the agents actually read. It follows Decant by a week — the local tool that parses Claude Code and Codex session logs into per-session cost and activity numbers — so the two releases sit on the same axis from opposite ends: Decant reads the logs on the developer's machine, Dosu now ingests them as a knowledge source. The monthly Drop format continues, with July's removing the waitlist and simplifying Knowledge Cache tools.
Dosu started by maintaining repository knowledge and is now positioning agent output as an input to it. That closes a loop: agents read the docs Dosu maintains, and their sessions become material Dosu learns from. The Drops also show a steady flattening of setup friction — waitlist removed, libraries and agents overhauled, configuration moved into chat — which is the pattern of a product trying to shorten time-to-value rather than widen its feature surface.
Expect the log ingestion and Decant's cost data to converge into one view of what agents cost against the maintenance work Dosu absorbs, which the August Drop's impact reporting now partially supplies.
The .NET and Python packages ship on a steady cadence, but the contents are servicing: SDK and package version bumps, CVE-driven dependency updates, CodeQL suppressions, and HTTP hardening such as disabling automatic redirects in the web plugins. The genuinely functional changes are narrow — a Gemini connector now honoring the configured function choice behavior, an MCP tool approval callback for Azure AI agents shipped as a breaking change, and MCP tools with colliding normalized names being skipped. The latest .NET release removes migrated vector-store providers outright, leaving redirect READMEs behind.
The centre of gravity is moving out of this repository. Vector store providers have migrated to CommunityToolkit packages and their originals are now deleted rather than deprecated, with samples following them across. What remains is maintenance plus the occasional MCP fix, which suggests the agent work that would once have landed here is happening in a different codebase. For teams with Semantic Kernel in production, the signal to read is the removals: each one is a dependency that now resolves somewhere else.
Expect the cadence to continue as security and dependency servicing with occasional MCP fixes, and for migration tooling or documentation pointing at Microsoft Agent Framework to grow faster than any new capability in Semantic Kernel itself.
Other ai-assistants products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either Dosu or Semantic Kernel.
Perplexity is selling access to other people's models, and now repricing them weekly.
Format coverage still outruns hardening — three corrective releases in five days
Copilot ships a model a week; now enterprises get switches for the plugins underneath
ONNX Runtime is dismantling itself into a core plus detachable accelerator plug-ins, CUDA included.
Transformers is becoming a dispatch layer over optimized kernels, and the patches now track vLLM's release calendar.
Alhena is slicing one benchmark study into a month of posts, one finding each.
See all Dosu alternatives → · See all Semantic Kernel alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Dosu is currently shipping more aggressively (velocity 6.3 vs 5.0), with 1 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. Dosu is currently shipping more aggressively (velocity 6.3 vs 5.0), with 1 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other ai-assistants products to evaluate alongside.
Top Dosu alternatives in ai-assistants are ranked by recent ship velocity. Browse the "Dosu alternatives" section above for the current picks, or visit /alternatives/dosu for the full list with editorial commentary on each.
Top Semantic Kernel alternatives in ai-assistants are ranked by recent ship velocity. Browse the "Semantic Kernel alternatives" section above for the current picks, or visit /alternatives/semantic-kernel for the full list with editorial commentary on each.